Laserfiche WebLink
referred to as "out -of- State" arrests). <br />Data on arrests came from State RAP <br />sheets and FBI RAP sheets. RAP <br />sheets (Records of Arrest and Prose- <br />cution) are law enforcement records <br />intended to document a person's entire <br />adult criminal history, including every <br />arrest, prosecution and adjudication for <br />a felony or serious misdemeanor <br />offense. Arrests, prosecutions and <br />adjudications for minor traffic offenses, <br />public drunkenness, and other petty <br />crimes are not as fully recorded as <br />those for serious crimes. The "percent <br />rearrested" is calculated by dividing the <br />number rearrested by the number <br />released from prison in 1994. <br />All measures of recidivism based on <br />criminal records are subject to two <br />types of errors. Type 1 errors arise <br />when the arrest or the conviction in the <br />released prisoner's record is for a <br />crime that person did not commit. <br />Type 2 errors arise when the released <br />prisoner commits a crime but he is not <br />arrested for it, or, even if he is, the <br />arrest does not result in his conviction. <br />Some amount of type 1 and type 2 <br />error is inevitable, however recidivism <br />is measured. But that does not mean <br />that all recidivism measures are equally <br />suitable, no matter the purpose they <br />are intended to serve. The main <br />purpose of this recidivism study was to <br />document the percentage of sex <br />offenders who continued their involve- <br />ment in various types of crime after <br />their release from prison in 1994. The <br />more suitable measure for that is the <br />one with the fewest type 2 errors: the <br />one, in other words, less prone to <br />saying someone is not committing <br />crimes when he actually is. Between <br />rearrest and reconviction as the recidi- <br />vism measure, the one less likely to <br />make that type of error is rearrest. <br />One reason is that the rigorous <br />standard used to convict someone — <br />"proof beyond a reasonable doubt" — <br />makes it certain that guilty persons will <br />sometimes go free. Another reason is <br />record keeping: the justice system <br />does better at recording arrests than <br />convictions in RAP sheets. For such <br />reasons, this study uses rearrest more <br />often than reconviction as the measure <br />of recidivism. <br />Rearrest forms a conservative meas- <br />ure of reoffending because many <br />crimes do not result in arrest. Not all <br />types of crime are alike in this regard. <br />Crimes committed in nonpublic places <br />(such as in the victim's home) by one <br />family member against another (such <br />as by the husband against his wife, or <br />by the father against his own child) are <br />a type that is less likely than many <br />other types to be reported to police <br />and, consequently, less likely to result <br />in arrest. Sex crimes, particularly those <br />against children, are a specific <br />example of this type. While some sex <br />offenders in this study probably com- <br />mitted a new sex crime after their <br />release and were not arrested or con- <br />victed, the study cannot say how many. <br />As mentioned above, one reason why <br />sex offenders are not arrested is that <br />no one calls the police. Results from <br />the National Crime Victimization Survey <br />indicate that the offenses of <br />rape /sexual assault are the least likely <br />crimes to be reported to the police. <br />(See Reporting Crime to the Police, <br />1993 -2000, March 2003, <http: / /www. <br />ojp. usdoj /bjs /abstract/rcp00.htm >.) <br />Reconviction Except where stated <br />otherwise, this recidivism measure <br />pertains to State and Federal convic- <br />tions in any State (not just convictions <br />in the State that released them) in the <br />three years following release. Informa- <br />tion on convictions came from State <br />and FBI RAP sheets. RAP sheets are <br />intended to document every conviction <br />for a felony or serious misdemeanor, <br />but not every conviction for a minor <br />offense. "Percent reconvicted" is calcu- <br />lated by dividing the number recon- <br />victed by the number released from <br />prison in 1994. (It is not calculated by <br />dividing the number reconvicted by the <br />number rearrested.) <br />6 Recidivism of Sex Offenders Released from Prison in 1994 <br />Return to prison Two recidivism <br />measures are returned to prison — <br />with a new sentence <br />with or without a new sentence. <br />Recidivism defined as Returned to <br />prison with a new sentence pertains <br />exclusively to sex offenders who, within <br />3 years following release, were recon- <br />victed for any new crime in any State <br />following their release and received a <br />new prison sentence for the new crime. <br />Recidivism defined as Returned to <br />prison with or without a new sentence <br />includes resentenced offenders plus <br />any who were returned to prison within <br />3 years because they had violated a <br />technical condition of their release. <br />Technical violations include things such <br />as failing a drug test, missing an <br />appointment with their parole officer, or <br />being arrested for a new crime. Offend- <br />ers returning to prison for such viola- <br />tions are sometimes referred to as <br />"technical violators." <br />Prisons should not be confused with <br />jails. A prison is a State or Federal <br />correctional facility reserved for <br />convicted persons with relatively long <br />sentences (generally over a year). <br />A jail is a local correctional facility for <br />convicted persons with short sentences <br />or for persons awaiting trial. Returns to <br />prison refer to any prison, not neces- <br />sarily the same prison that released the <br />offender in 1994. <br />The "percent returned to prison with a <br />new sentence" is calculated by dividing <br />the number returned to prison with a <br />new sentence by the number released <br />from prison in 1994. The "percent <br />returned to prison with or without a new <br />sentence " is calculated by dividing the <br />number returned to prison with or <br />without a new sentence by the number <br />released from prison in 1994. <br />Data on returns with a new sentence <br />are based on State and FBI RAP <br />sheets. Data on returns with or without <br />a new sentence are based on State <br />and FBI RAP sheets plus prison <br />records. <br />